High-Throughput Real-Time Quantitative · High-Throughput Real-Time Quantitative Reverse Transcription PCR 15.8.2 Supplement 73 Current Protocols in Molecular Biology amount of RNA

UNIT 15.8High-Throughput Real-Time QuantitativeReverse Transcription PCR

This unit describes the use of real-time quantitative PCR (QPCR) for high-throughputanalysis of RNA expression. The topics covered include: the standard curve method(see Basic Protocol 1); production and quantification of RNA standards (see SupportProtocol 1); an efficiency-corrected �Ct (cycle time, also called cycle threshold orcrossing point) method (see Basic Protocol 2); the comparative cycle time, or ��Ctmethod (see Alternate Protocol); and design and validation of QPCR primers and probesfor both SYBR Green– and TaqMan-based assays (see Support Protocol 2). While theunit describes the use of the Applied Biosystems 7900HT (high-throughput, 384-well)instrument, the protocols may be utilized for any real-time PCR instrument. The high-throughput design allows analysis of the levels of transcripts from a number of genes ofinterest (GOIs) at one time by using the appropriate primer set for each gene. (Withinthis unit, the term GOI will refer to the actual gene of interest as well as its RNA productor cDNA copy.)

Absolute quantification means that the absolute copy number of the GOI is measured.Relative quantification means that a quantitative difference in copy number betweentwo samples, experimental and control, is measured by normalizing both samples to anendogenous reference.

Because of the simplicity of the mathematical application, the relative standard curvemethod (Basic Protocol 1) is the most basic and straightforward QPCR assay describedin the unit. In this method, standard curves are constructed for all of the GOIs from whichRNA expression is being measured, and linear regression analysis is applied to interpolateunknown sample values. The standard curve assay may be performed even if the PCRamplification efficiencies of the primer sets (as determined by the template dilution assayin Support Protocol 2) are not equal, since correction for unequal efficiencies is intrinsicto the linear regression formula. One drawback of the standard curve method is thatstandard curves must be run for each of the primer sets on an assay plate. This resultsin less space on the plate for the unknown samples, and requires the use of additionalreagents. This use of resources is particularly excessive when the PCR amplificationefficiencies of the primer sets have been determined to be essentially 100% and relativefold-change is the preferred outcome of the measurements. In such cases, the ��Ctmethod (see Alternate Protocol) should be employed instead. Another limitation is thatunless the levels of all of the GOIs in the cDNAs used to construct the standard curvesare known, the relative concentration of one GOI cannot be compared to that of anotherGOI. If comparison between the levels of different GOIs (without the knowledge of therelative level of the transcripts in the standards) is desired, the efficiency-corrected �Ctmethod (see Basic Protocol 2) should be applied. Alternatively, absolute standards canbe generated (Support Protocol 1).

The efficiency-corrected �Ct method builds upon the relative standard curve methodby incorporating PCR efficiency (E) into the quantity calculations. The standard curveslopes are used to calculate PCR efficiency according to the relationship E = 10(–1/slope).The efficiency has a maximum value of 2 for perfect doubling of the PCR template(see Basic Protocol 2 for an in-depth explanation of E). Note that in this method thestandard curve is used only to determine slope and not to interpolate the RNA values ofthe unknown samples. The efficiency correction is then applied to determine the relative

Contributed by Angie L. Bookout, Carolyn L. Cummins, Martha F. Kramer, Jean M. Pesola, andDavid J. MangelsdorfCurrent Protocols in Molecular Biology (2006) 15.8.1-15.8.28Copyright C© 2006 by John Wiley & Sons, Inc

The PolymeraseChain Reaction

15.8.1

Supplement 73

High-ThroughputReal-Time

QuantitativeReverse

Transcription PCR

15.8.2

Supplement 73 Current Protocols in Molecular Biology

amount of RNA using the measured Ct values for the test samples, and is particularlyimportant when comparing the levels of different RNAs whose standard-curve slopesdeviate from each other by greater than ±0.1. This application is useful for comparingthe expression profiles of many different RNAs, e.g., those belonging to gene familiesor related biological pathways. The sample-space and reagent-use limitations mentionedfor the standard curve method also apply here.

The ��Ct method is the method of choice when the desired output is “fold-change,”because standards are not necessary, thus saving both reagents and space on the re-action plate. However, this method requires that the amplification efficiencies of theprimer/probe sets be 100%. If the amplification efficiencies are suboptimal, but theprimers generate a single product as determined by melting curve analysis (SupportProtocol 2), Basic Protocol 1 should be used to determine fold-changes.

Chapter 4 describes the isolation of RNA from several sources and UNIT 15.5 detailsthe traditional procedure for reverse transcription of RNA into complementary DNA(cDNA). For QPCR, it is recommended that total RNA be resupended in diethylpyrocar-bonate (DEPC)–treated water or an equivalent nuclease-free buffer that does not containEDTA. The RNA should be treated with DNase and then reverse transcribed using 0.08µg/µl (final concentration) random hexamer or nonamer primers. Purified messengerRNA—i.e., poly(A)+ RNA (UNIT 4.5) or RNA that has been reverse transcribed usingoligo(dT) or gene-specific reverse primers—may also be successfully used in the assay.See Commentary for more detailed information.

Following the assay, the resulting raw data are analyzed using second-party software,usually Microsoft Excel or equivalent. The data analyses are dependent on the type ofassay performed, and are outlined in detail as part of each protocol.

NOTE: General precautions for working with RNA are described in UNIT 4.1 and otherChapter 4 units, and general precautions necessary for PCR are described in UNIT 15.1.In particular, the use of molecular-biology-grade water, RNase/DNase/nucleic acid–free tubes, aerosol-barrier pipet tips, and dedicated pipettors of all types (i.e., pipettorsused only for RNA or PCR applications, which are kept out of areas used for plasmidor genomic DNA work) is strongly recommended for all steps in this unit. If dedicatedpipettors are not available, the available pipettors should be thoroughly cleaned to removenucleases and potential contaminants such as plasmid or genomic DNA. In addition, theuse of gloves is required, since even a small amount of any contaminant can greatlyimpact the results of the assay.

STRATEGIC PLANNING

To begin performing a QPCR assay, design and validation of the appropriate primers andprobes must first be completed (refer to Support Protocol 2). Second, the appropriateassay is selected based on the goal of the experiment and the desired data output (refer toBasic Protocols 1 and 2, and the Alternate Protocol, for guidelines used in making thisdetermination). The chosen assay is then performed and the data are analyzed.

BASICPROTOCOL 1

STANDARD CURVE METHOD FOR RELATIVE QUANTIFICATION

The relative standard curve method is used for determining the level of a gene of interest(GOI) relative to an endogenous reference RNA, and for calculating relative fold-changesof a GOI between experimental samples. The assay is useful for determining an “expres-sion profile” of a single GOI within a group of samples. A dilution series of standardcDNA samples is constructed for the GOI and reference gene, and linear regressionanalysis is applied. The formulas resulting from the standard curves are used to inter-polate the GOI and reference-gene quantities in the unknown samples. An endogenous


15.8.3

Current Protocols in Molecular Biology Supplement 73

reference gene, often a housekeeping gene, is used as a control to normalize the amountof input template for each sample (see Commentary for parameters used in choosingthe appropriate reference gene). The data are expressed as “normalized RNA level” inarbitrary units. The type of nucleic acid standard chosen depends upon the nature of theunknown samples and is discussed in the Commentary.

The relative standard curve assay does not require that the amount of the GOI or referenceRNA in the standards be known. It depends on the linear regression formula produced byplotting the Ct versus the log nanogram (log ng) of input standard total RNA. It should benoted that cDNA concentrations are not typically determined following reverse transcrip-tion of the RNA. Here, quantity refers to total RNA input prior to reverse transcription.The standard curve–plotting function is available in most instrument software. If it isnot, graphing software may be used instead. Since the input ng values refer to the inputamount of total RNA, and not to a known amount of target molecules, the numbersgenerated are simply arbitrary and may not be compared with the numbers calculated fora different GOI. If comparison of the relative RNA levels between different RNA targetsis desired, refer to the efficiency-corrected �Ct method (Basic Protocol 2) or absolutequantification (Support Protocol 1). The standard curve method should be used instead ofthe ��Ct method (Alternate Protocol) to find fold-changes between samples when theamplification efficiencies of the primer sets are not 100% as determined by the templatedilution assay (Support Protocol 2). If absolute levels of transcript are desired, syntheticRNA standards may be prepared through incorporation of radionucleotides (see SupportProtocol 1) and used in standard curve analysis.

Materials

20 ng/µl experimental cDNA samples (concentration based on RNA input forcDNA synthesis; see UNIT 15.5)

Dilution series of standard cDNAs (see recipe) or dilution series of [35S]RNAstandards (see Support Protocol 1)

No-template control sample (NTC; prepared at the same time as the cDNA samplesusing molecular-biology-grade water instead of RNA; see Commentary)

No-reverse-transcriptase control samples (–RT; prepared at the same time as thecDNA samples using molecular-biology-grade water instead of reversetranscriptase; see Commentary)

2× SYBR Green or TaqMan mix containing ROX (Applied Biosystems, Bio-Rad,Invitrogen, Sigma, or see recipe for 2× SYBR Green mix)

Primer mixes, 1.25 µM each forward and reverse primer (see recipe and SupportProtocol 2), for each reference gene and GOI to be tested

5 µM TaqMan probe (for TaqMan protocol only; see recipe and Support Protocol 2)Molecular-biology-grade water (nucleic acid and nuclease free)8-tube PCR tube strips (optional, but recommended; can be of low quality since

they will only be used for mixing reaction components; ISC Bioexpress)96-well PCR tube racks (optional, but recommended; ISC Bioexpress)Digital multichannel pipettor, 8- or 12-channel, 5- to 100-µl capacity

(recommended)Centrifuge with swinging-bucket rotor and microtiter plate carriers384-well optical reaction plates (Applied Biosystems)Optical adhesive covers (Applied Biosystems)Real-time thermal cycler: e.g., Applied Biosystems 7900HTMicrosoft Excel or spreadsheet program with equivalent statistical features

Set up plates1. Plan the plate arrangement according to the number of samples and primer sets to

be assayed (reference gene plus GOIs). For each primer set, include the standards,


QuantitativeReverse

Transcription PCR

15.8.4


Figure 15.8.1 Typical plate setup for the standard curve and efficiency-corrected �Ct assays. (A) Organization ofPCR tube strips on a 96-well PCR rack when preparing for any QPCR assay. It is advisable to premix the cDNAtemplates with the primer master mixes in tube strips before putting them into the reaction plate. This practicedecreases the variability between replicate wells. A typical cDNA/primer mix setup is shown in relation to the final384-well reaction plate (B). The plate arrangement shown represents an experiment in which 20 samples takenfrom experimental animals will be assayed for levels of one endogenous control RNA and three GOIs. The samplesare plated in triplicate for each of the RNAs assayed, and a standard curve is required for each of the RNAs tobe measured (the row before the unknown samples contains the standard samples, whose amounts are shown inng total RNA). Abbreviations: GOI, gene of interest; NTC, no-template control; Ref, endogenous control gene; –RT,no-reverse-transcriptase control.

the NTC, and the –RT control as part of the sample group. See Figure 15.8.1 for anexample of a typical plate setup.

All steps may be performed at room temperature if a vendor-supplied SYBR Green orTaqMan mix is used. Most vendor-supplied 2× mixes are stable at room temperature fora number of hours.

If the instrument is equipped with a plate-loading robotic arm, several plates may beprepared at once and stacked into the robotic arm queue. Alternatively, plates may bemade several days in advance and stored at 4◦C until ready to run, without a decline inthe performance of the PCR enzyme. Advance preparation of plates is not recommendedif SYBR Green or TaqMan mixes are prepared in the laboratory, since the home-mademixes lack additives present in the commercial mixes that confer stability to the reactioncomponents.


15.8.5


Table 15.8.1 Master Mixes for the Standard Curve and Efficiency-Corrected �Ct Assays

ComponentsFinalconcentration

Volume perwell

Volume for eachcDNA/primer mix (persample in triplicate + 1extra)

Master mix (no. ofsamples + 8 standardsa

+ 1 extra)

SYBR Green assay

2× SYBR Green mix 1× 5 µl 20 µl

1:1 primer mix (1.25µM each)

150 nM eachprimer

1.2 µl 4.8 µl

Template cDNA 10-25 ngb 1.25 µl 5 µl —

H2O N/A to 10 µl to 40 µl

Total volume 10 µl 40 µl Aliquot 35 µl into eachtube containing cDNA

TaqMan assay

2× TaqMan mix 1× 5 µl 20 µl


300 nM eachprimer

2.4 µl 9.6 µl

5 µM TaqMan probe 250 nM 0.5 µl 2.0 µl

Template cDNA 10-25 ngb 1.25 µl 5 µl —



aFor this assay, the no-template control (NTC) and no-reverse-transcriptase (–RT) control are included as part of the standard sample set.bRecommended amount of template for detection of both high and low levels of GOIs. If necessary, significantly less template may be used(picograms). Note that the cDNA template quantity is based upon the amount of total RNA input into the reverse transcription reaction.

2. Prepare primer master mixes according to Table 15.8.1, but without template cDNA.

This can be done in advance, and the tubes may be put on ice or stored at 4◦C for a fewhours before preparing the reaction plate.

The authors always use the same concentrations of PCR components because this greatlyincreases the high-throughput nature of the assay. Keeping the conditions of the mastermixes constant allows not only universal mix conditions but universal cycling conditions.If the initial primer set does not perform well, new primers are designed (also see SupportProtocol 2).

3. Place the appropriate number of 8-tube PCR strips into a 96-well PCR tube rack (seeFig. 15.8.1A). For convenience, use a different color tube strip for each differentprimer master mix that has been prepared. Label the side of each strip with the letterof the row of the reaction plate into which the samples will be placed.

Alternatively, cDNA/primer mixes may be made in 0.65-ml microcentrifuge tubes or 0.2-or 0.5-ml PCR tubes.

The cDNA may be put directly into the optical reaction plates followed by the primermaster mix. However, this is not recommended because it introduces a potential source ofexperimental error, since the assay replicates are not premixed, but pipetted individually.The end result may be lower assay precision.

4. Using a multichannel pipettor, put 5 µl of cDNA (experimental samples, standards,and controls) into the bottom of the appropriate tubes in the strips.


QuantitativeReverse

Transcription PCR

15.8.6


5. Being careful not to touch the cDNA inside the tubes, use a multichannel pipettor toplace a 35-µl aliquot of the appropriate primer master mix into each tube.

6. Cover the entire rack of tube strips with Parafilm and gently vortex to mix. Gentlytap or briefly centrifuge the PCR tube racks (2 to 3 min at 1700 × g, 4◦C or roomtemperature, in a swinging-bucket rotor with microtiter plate carriers) to get contentsto the bottoms of the tubes.

7. Using a multichannel pipettor, dispense 10 µl of each cDNA/primer mix into theappropriate three wells of the optical reaction plate to generate each sample intriplicate (as planned in step 1; see Fig. 15.8.1B).

If the multichannel pipettor has 8- or 12-channel dispensing capability, the triplicatescan be dispensed at the same time for different rows. Note that due to the spacing betweenrows of a 384-well reaction plate, every other row can be added at once (i.e., A, C, E, G,I, K, M, O, and then B, D, F, H, J, L, N, P), thus greatly minimizing the pipetting time.

8. Cover the plate with the optical adhesive cover and then briefly centrifuge the plateas above to get contents to the bottoms of the wells.

Perform real-time PCR9a. For real-time PCR: Transfer the plate to the real-time thermal cycler and run real-

time PCR using the following program (consult the instrument manual for specificinstructions):

1 cycle: 10 min 95◦C (activates the hot-start Taq DNA polymerase)40 cycles: 15 sec 95◦C (collect data throughout)

1 min 60◦C (collect data throughout).

9b. For melting (dissociation) curve analysis (for use with SYBR Green only): Add thesesteps following the 40 cycles of the thermal cycling program.

15 sec 95◦C15 sec 60◦C (collect data)Increase from 60◦ to 95◦C at a 2% temperature ramping rate (collect data)15 sec 95◦C (collect data).

See Support Protocol 2 for a description of the use of melting curve analyses.

Analyze data10. Analyze and export raw data (see instrument manual for detailed instructions about

document setup, baseline, and threshold settings).

Some instrument software applications contain a standard curve plotting feature. If thisfunction is not available, use Excel or another graphing program to plot Ct versus the lognanograms (ng) of input total RNA for each standard, and apply a best-fit line to generatethe linear regression formula y = mx + b, where y is Ct of the unknown sample, m is theslope, x is the quantity of the unknown sample (in log ng), and b is the y intercept for boththe reference gene and each GOI. Interpolate the unknown sample quantities using theresulting formulas.

The transformation of the fluorescence signal into Ct data, as well as methods for baselineand threshold settings, vary by instrument. The specific instrument manual should beconsulted.

In analyzing the raw data, it is important to adjust the cycle threshold (Ct) of the am-plification plot to within the geometric phase of amplification. This is critical for properanalysis because the geometric phase represents the point of the reaction at which Ct isquantitatively related to the amount of initial PCR template. Note that a Ct decrease of 1unit represents a two-fold increase in initial PCR template.


15.8.7


It is also important that the coefficient of determination, or R2value, for the linearregression formula be 0.99. If the R2 value is less than 0.99, this suggests that oneor more points of the standard curve are deviating significantly from the best-fit line. Inthis case, the accuracy of the data obtained from the linear regression formula of thisstandard curve may be compromised. The R2 value will never be above 0.99.

11. Import data into Microsoft Excel or equivalent spreadsheet program with statisticalfeatures.

12. For each of the three replicates of a sample, calculate the average quantity (avg)of target cDNA interpolated from the standard curve, the standard deviation ofthe average (stdev), and the coefficient of variation (CV) according to the formulaCV = stdev/avg.

13. Remove any outlier points (>17% CV). After removing the outlier point, recalculateavg, stdev, and CV.

Only one point per replicate may be removed.

A 17% CV correlates with the maximum allowable standard deviation that can distinguisha two-fold change with 99% confidence when samples are assayed in triplicate wells forboth the endogenous reference and the GOI. If a 95% confidence interval is acceptable,a 21.8% CV may be used as the threshold for removing outliers. On the other hand, theQ-test (a test for rejection of discordant data) may be used to determine outlier points.Refer to Shoemaker et al. (1974) for a more in-depth description of this test.

14. For each sample, normalize the GOI quantity to that of the reference gene for thesample according to the following equation. Use the recalculated values if outlierpoints were removed in step 13.

15. Calculate the standard deviation (SD) of the normalized value according to theequation:

16. Plot the resulting values as a bar graph of normalized value versus sample name orexperimental treatment group, with the error bars equal to the SD.

17. If desired, calculate fold-changes between samples by choosing a calibrator sample(usually vehicle-treated or wild-type control) and dividing all of the normalizedvalues from step 14 and the SD calculated in step 15 by the normalized value of thiscalibrator.

The resulting values are then expressed as fold-changes relative to the calibrator sample,which should now be equal to 1.

BASICPROTOCOL 2

EFFICIENCY-CORRECTED �Ct METHOD

The efficiency-corrected �Ct method is used for determining the relative amounts of dif-ferent GOIs that are normalized to an endogenous reference RNA (e.g., 18S cyclophilin).It may also be used to determine fold-changes of a specific RNA between samples, butmay be excessive if the desired output is only fold-change (see Basic Protocol 1 or Alter-nate Protocol). Data obtained from the efficiency-corrected �Ct method are expressed as“normalized RNA level” in arbitrary units, and the calculated levels may be compared tothose of other GOIs when the same threshold setting and assay chemistry are used (i.e.,SYBR Green or TaqMan chemistry). It should be noted that an assumption is made thatthe reverse transcription efficiency is equal for all RNA transcripts in a single sample


QuantitativeReverse

Transcription PCR

15.8.8


and for the same transcript between samples. In some cases, this may not be true (seePfaffl, 2004, for further discussion); therefore, it is recommended that the RNA extrac-tion method remain the same for all samples, and that all samples under study be reversetranscribed at the same time with the same reaction buffer.

The basis of quantitative PCR lies in the principle that for every additional thermocycle,a two-fold increase of template-specific product occurs. Several factors affect whether achange in one cycle truly represents a two-fold growth in product, in other words, whetherthe reaction is 100% efficient. To assess the reaction’s efficiency, linear regression analysisis applied to a standard cDNA dilution series, just as in Basic Protocol 1. The slope ofthe resulting standard curve is used as a measure of PCR efficiency (E) according to theequation E = 10(–1/slope). Note that different GOIs may produce different E values.

A slope of −3.3 produces an E value of 2, indicating that a perfect doubling of thetemplate has occurred. Calculated E values of less than 2 imply that the template hasnot been perfectly doubled. Template, primer, and probe quality and quantity, samplecomplexity, and pipet calibration and buffer conditions like MgCl2, salt, additives, anddeoxynucleotide concentrations, all contribute to this efficiency (Pfaffl, 2004; see UNITS

15.1 & 15.5 for further details on these parameters). Given optimal buffer conditions andadequate primer, probe, and sample qualities, slight fluctuations in efficiency may stillbe observed between primer sets run on the same plate or between assay plates, evenif they have been assembled and run on the same day by the same user. Efficiencycorrection is a means to account for inter- or intra-assay variability that is attributable tothe aforementioned parameters.

The materials and setup of the assay are the same as in the relative standard curve method(see Basic Protocol 1), except that the linear regression formula produced by plottingthe Ct versus the log ng of input standard total RNA is used only to determine PCRefficiency. This computed efficiency is then used to calculate the RNA levels (in arbitraryunits) of the GOI and the endogenous control genes. The GOI RNA level in each sampleis then expressed as a ratio relative to the endogenous control RNA level in that sample.Because the data are dependent upon Ct values and not a standard curve, the resultingvalues may be compared to those of another RNA.

1. Set up and run assay as described in Basic Protocol 1 (steps 1 to 9a). Refer to Figure15.8.1 for an example of a typical plate setup and to Table 15.8.1 for master mixcomponents.

2. Analyze and export raw data (see Basic Protocol 1, step 10), and then import intoMicrosoft Excel or equivalent spreadsheet program.

The threshold values for all RNAs measured (including the endogenous reference) mustbe the same. It is important to determine a suitable threshold within the geometric phaseof the amplification plots for all RNA transcripts to be compared.

3. Calculate PCR efficiency, E = 10(–1/slope), for the endogenous control RNA and eachGOI from the slopes of their corresponding standard curves.

4. Calculate the quantity of the endogenous control RNA and each GOI from their Ctvalues according to the formula quantity = E−Ct.

When the efficiency is 100% (i.e., slope = –3.3 and E = 2), the equation becomes quantity= 2−Ct. This serves as the basis for the calculation performed in ��Ct method (AlternateProtocol).

5. For each of the three replicates of a sample, calculate the average quantity (avg),the standard deviation of the average (stdev), and the coefficient of variation (CV),where CV = stdev/avg.


15.8.9


6. Remove outliers, normalize the GOI, calculate the SD, and plot the results (see BasicProtocol 1, steps 13 to 17).

ALTERNATEPROTOCOL

COMPARATIVE OR ��Ct METHOD

The comparative Ct or ��Ct method is used for measuring the fold-changes in expressionof a particular RNA transcript between experimental samples. Typically, this assay isused when investigating gene-expression differences between wild-type and knockoutor transgenic animals, or between vehicle-control and drug-treated samples. The resultsare then expressed as “fold-changes” relative to a calibrator, such as an untreated orwild-type sample. The ��Ct method is only applicable when the primer sets for boththe GOI and the endogenous reference gene have been shown to give perfect standardcurve slopes (slopes = −3.3 ± 0.1 with R2 = 0.99) as assayed in Support Protocol 2.

1. Set up assay as described in Basic Protocol 1 (steps 1 to 9a), except refer to Figure15.8.2 for an example of a typical plate setup and to Table 15.8.2 for master mixcomponents.

Standard cDNA samples are not needed in this method.

2. Analyze and export raw data (see Basic Protocol 1, step 10) and then import intoMicrosoft Excel or equivalent spreadsheet program.

3. For each of the three replicates of a sample, calculate the average (avg) cycle time(Ct) and then calculate the standard deviation (stdev).

4. Remove any outlier wells from the averaged Ct values (>0.3 stdev).

Only one point per replicate may be removed.

%CV may not be used, due to the logarithmic nature of both the Ct avg and Ct stdev.Instead, stdev must be used, where a stdev of 0.3 correlates with the maximum allowablestandard deviation that can distinguish a two-fold change with 99% confidence; 0.4 stdevmay be used for a 95% confidence interval.

Figure 15.8.2 Typical plate setup for the ��Ct method. The plate arrangement shown represents an experiment inwhich 20 samples taken from experimental animals will be assayed for one endogenous control gene and four GOIs. Thesamples are plated in triplicate for each of the RNAs assayed. Abbreviations: GOI, gene of interest; NTC, no-templatecontrol; Ref, endogenous control gene.


QuantitativeReverse

Transcription PCR

15.8.10


Table 15.8.2 Master Mixes for the ��Ct Assay


Volume perwell


Master mix (no. ofsamples + 1 NTC + 1extra)

SYBR Green assay

2× SYBR Green mix 1× 5 µl 20 µl


150 nM eachprimer

1.2 µl 4.8 µl

Template cDNA 10-25 nga 1.25 µl 5 µl —



TaqMan assay

2× TaqMan mix 1× 5 µl 20 µl


300 nM eachprimer

2.4 µl 9.6 µl

5 µM TaqMan probe 250 nM 0.5 µl 2.0 µl

Template cDNA 10-25 nga 1.25 µl 5 µl —



aRecommended amount of template for detection of both high and low levels of GOIs. If necessary, significantly less template may be used(picograms). Note that the cDNA template quantity is based upon the amount of total RNA input into the reverse transcription reaction.

5. For each sample, normalize the GOI Ct values to those of the reference gene for thesame sample according to the equation:

Calculate the standard deviation of �Ct (stdev�Ct) as:

6. Choose a calibrator.

This will be the sample, tissue, gene, or control group to which the others will be compared.

7. Find the ��Ct, or calibrated value, for each sample, according to the equation:

The stdev��Ct will be the same as stdev�Ct, since the calibrator is arbitrarily set to be aconstant.

8. Find the fold-change for each sample relative to the calibrator according to theequation:

For the sample that is chosen as the calibrator, the ��Ct = 0 and therefore the fold-change = 2(−��Ct)= 1.


15.8.11


9. Plot the resulting fold-changes on a bar graph of fold-change versus sample name orexperimental treatment group. Determine the measure of experimental error as:

SUPPORTPROTOCOL 1

GENERATION OF RNA STANDARDS FOR ABSOLUTE QUANTIFICATIONBY REVERSE TRANSCRIPTION PCR

Absolute quantification of RNA molecules in unknown samples by reverse transcriptionPCR (RT-PCR) requires knowledge of the copy number of specific RNA molecules.These can be subjected to reverse transcription and PCR in the same manner as theexperimental samples, thus accounting for the reaction efficiencies of both procedures.This protocol describes the production and quantification of synthetic RNA standardsfor use in Basic Protocol 1 to determine absolute amounts, instead of relative levels,of a GOI. RNA standards are produced via in vitro transcription in the presence oftrace amounts of an 35S-labeled ribonucleoside triphosphate (rNTP), which permitsaccurate quantification of transcripts by measurement of 35S incorporation. This protocolencompasses template construction, in vitro transcription, DNase treatment, monitoringthe efficiency of the reactions by agarose gel electrophoresis, and determination of yield.To determine yield, synthesized RNA can be quantitatively precipitated and purified awayfrom unincorporated nucleotides by application to filters followed by trichloroacetic acidwashes. This protocol describes a batch precipitation/washing method. For an individualfiltration method, see UNIT 3.4. The preparation and use of RNA standards in quantitativeRT-PCR assays have been described by Kramer and Coen (1995).

Materials

cDNA or DNA fragment containing target sequenceVector containing T7, T3, or SP6 RNA polymerase promoterAppropriate restriction enzyme (UNIT 3.1) for linearizing plasmid25:24:1 (v/v/v) phenol/chloroform/isoamyl alcohol49:1 (v/v) chloroform/isoamyl alcohol3 M sodium acetate (APPENDIX 2)100% ethanolNuclease-free water0.5 µg/ml sheared salmon sperm DNA1 M Tris·Cl, pH 7.5 (APPENDIX 2)1 M MgCl2 (APPENDIX 2)1 M DTT (APPENDIX 2)500 µM 4NTP mix: 500 µM each ATP, CTP, GTP, and UTP600 Ci/mmol (10 mCi/ml) [α-35S]CTP or UTPBovine serum albumin (BSA)Spermidine (for SP6 only)T7, T3, or SP6 RNA polymerase (UNIT 3.8)1 U/µl RNase-free DNase I (e.g., Promega)RNeasy Mini Kit (Qiagen) or equivalent10% trichloroacetic acid (TCA; see recipe), ice cold100% methanol, ice coldUniversal scintillation cocktail (preferably biodegradable, e.g., Ecoscint A,

National Diagnostics)

1.5-ml screw-cap microcentrifuge tubesDE81 paper (Whatman)Filter paper


QuantitativeReverse

Transcription PCR

15.8.12


Glass fiber filters (Whatman GF/C 24-mm discs)Transparent plastic wrapForcepsHeat lamp (optional)250-ml glass or metal beakerLiquid scintillation counter and vials

Additional reagents and equipment for subcloning (UNIT 3.16), plasmid minipreps(UNIT 1.6), digestion with restriction endonucleases (UNIT 3.1), agarose gelelectrophoresis (UNIT 2.5A), purification of DNA (UNIT 2.1A), phage RNApolymerase reactions (UNITS 3.4 & 3.8), agarose-formaldehyde or glyoxal gelelectrophoresis (UNIT 4.9), and drying and imaging of gels (APPENDIX 3A)

CAUTION: To minimize the risk of radioactive contamination, use screw-cap microcen-trifuge tubes with cap gaskets and filtered pipet tips for all manipulations of radioactivesolutions. Wear gloves and dispose of all 35S-contaminated material properly. See AP-

PENDIX 1F for more details on handling radioactivity.

CAUTION: Phenol, chloroform, and trichloroacetic acid are hazardous (see APPENDIX 1H).

NOTE: For all procedures involving RNA, use reagents and solutions that are freeof contaminating RNases, DNases, and nucleic acids, and follow other guidelines forhandling RNA (UNIT 4.1).

Construct and prepare template for in vitro transcription1. Construct a transcription plasmid by subcloning (UNIT 3.16), placing the sequence

contained in the RNA of interest downstream of the promoter for one of the threebacteriophage RNA polymerases (T7, T3, or SP6). Isolate the plasmid using aminiprep (UNIT 1.6), which will provide sufficient amounts of plasmid DNA for thisprotocol.

There are numerous commercially available vectors that contain phage polymerase pro-moters on one or both sides of a multiple cloning site (see Table 2.10.1). Additionalconsiderations for the construction of this plasmid are discussed in Critical Parametersand Troubleshooting.

2. Linearize 10 µg of plasmid with a restriction enzyme (UNIT 3.1) that will generate atemplate for run-off transcription. Run 5% of the linearization reaction volume onan agarose gel (UNIT 2.5A) alongside a control sample of uncut plasmid to confirm thatthe digestion is complete and yields the expected product size(s).

Long transcription products from incompletely digested plasmids can lead to inaccuraciesin quantification and extraneous products from RT-PCR.

3. Purify the linearized template from the reaction mixture using two or three organicextractions with 25:24:1 (v/v/v) phenol/chloroform/isoamyl alcohol (UNIT 2.1A), oneextraction with 49:1 (v/v) chloroform/isoamyl alcohol, and finally ethanol precipita-tion (UNIT 2.1A) in the presence of sodium acetate. Resuspend DNA in nuclease-freewater.

Acceptable results may also be obtained by using an enzymatic reaction clean-up kit (suchas the QIAquick PCR Purification Kit, Qiagen) or by performing gel extraction (UNIT 2.6).

If RNase contamination is a problem, the sample can be treated with 50 to 100 µg/mlproteinase K to destroy the RNases prior to purification.

The concentration of linearized template does not need to be determined. The authorsassume that most of the 10 µg used in the digest are recovered.

Synthesize RNA4. For each transcription reaction, prepare four 1.5-ml screw-cap microcentrifuge tubes

each containing 24 µl of 0.5 µg/ml sheared salmon sperm DNA: one pair to measure


15.8.13


input 35S in duplicate and one pair to measure 35S incorporation in duplicate. Alsoprepare one single tube as a control with no other additions. Set these tubes aside onice.

The salmon sperm DNA will act as a carrier for the synthesized RNA during the acidprecipitation step.

5. In another 1.5-ml screw-cap microcentrifuge tube, set up a 100-µl transcriptionreaction (minus the enzyme) as follows:

40 mM Tris·Cl, pH 7.510 mM MgCl25 mM DTT500 µM 4NTP mix1 µl 10 mCi/ml (600 Ci/mmol) [α-35S]CTP2 to 5 µg DNA template (from step 3)5 µg BSA1 mM spermidine (for SP6 only).

If the buffer contains spermidine, reaction components other than the enzyme should beat room temperature to avoid precipitation of the DNA.

If multiple reactions are being performed, a master mix containing everything but thetemplate should be prepared.

In vitro transcription kits can be used for convenience, but they will increase the cost.

6. Mix reaction components well. Transfer 1 µl of the reaction to each of two tubes ofsalmon sperm DNA from step 4 (input 35S samples). Set aside on ice.

7. Add 25 to 50 U RNA polymerase (T7, T3, or SP6) to the reaction mix (reactionvolume should now be 98 µl). Incubate 1 to 2 hr at 37◦C.

8. Transfer 5 µl of the reaction mixture to a separate tube and store on ice.

This is the pre-DNase sample that will be assessed by gel electrophoresis.

9. Add 5 µl (5 U) RNase-free DNase I or similar enzyme and incubate at 37◦C for30 min.

Transcription buffers typically contain sufficient magnesium ion concentrations (≥6 mM)to support DNase activity.

10. Purify the synthetic RNA using the RNeasy Mini Kit or similar product. Perform thefinal elution twice to maximize yield.

RNA can also be purified using organic extraction and isopropanol precipitation (UNIT

4.1); however, this removes less of the unincorporated rNTPs and is especially not recom-mended for difficult templates where yields are low (e.g., GC-rich sequences).

11. Elute or resuspend RNA in 93 µl nuclease-free water.

This volume is equal to the volume after step 8, thus making the gel samples from steps 8and 13 directly comparable.

12. Transfer 1 µl of the purified RNA to each of two tubes of salmon sperm DNA fromstep 4 (incorporated 35S samples). Set aside on ice.

13. Transfer 5 µl of the purified RNA to a separate tube and store on ice.

This is the final RNA sample that will be assessed by gel electrophoresis.

14. Store the remainder of the RNA at −70◦C in multiple aliquots to avoid repeatedfreeze/thaw cycles.


QuantitativeReverse

Transcription PCR

15.8.14


The size of the aliquots will depend on their intended use. It is convenient to prepare eachaliquot with enough RNA for two or three samples.

Use low-retention/nonstick microcentrifuge tubes, if possible, because these prevent theadherence of RNA to the tube walls over time.

Confirm in vitro transcription product by gel electrophoresis15. Electrophorese the pre-DNase (step 8) and final (step 13) samples through a formalde-

hyde or glyoxal agarose gel (see UNIT 4.9) of an appropriate concentration for the lengthof the RNA. Visualize synthesized products by staining with ethidium bromide (UNIT

4.9).

Make the gel as thin as possible to hasten drying later. Verify that any plasmid DNApresent in the pre-DNase sample (usually observed as a low-mobility band) is absentfrom the final sample. The successful removal of DNA should also be confirmed at theRT-PCR stage with a control lacking reverse transcriptase.

16. To dry the gel, first trim the gel to contain only relevant lanes. Stack one sheet ofWhatman DE81 paper on top of five sheets of filter paper, all cut several inches largerthan the gel. Center the gel on top of the DE81 paper and cover stack with a sheetof plastic wrap. Dry as for a polyacrylamide DNA gel (APPENDIX 3A), increasing thedrying time as needed for the thickness of the agarose gel.

The DE81 paper should retain most of the unincorporated [α-35S]rNTP while the blottingpaper will absorb excess moisture (and radioactivity).

CAUTION: Radioactive contamination of the gel dryer may occur. Be sure to conductrequired surveys and decontamination as directed by the institutional radiation safetycommittee (also see APPENDIX 3F).

17. Visualize labeled transcripts by autoradiography or by exposing to a phosphor screen(APPENDIX 3A).

Confirm that there is only one major radiolabeled product of the expected size (seeCommentary). Also confirm that the amount of synthetic RNA represented by the pre-DNase sample was efficiently recovered in the final sample. If the gel is not run too far, itwill be possible to visualize the amount of unincorporated [α-35S]rNTP removed by thepurification procedure.

Quantify 35S incorporation and RNA yield18. Label glass fiber filters by cutting varied notches along the perimeter, and lay them

out on a sheet of plastic wrap with forceps.

Prepare four filters for the single tube of control salmon DNA (step 4), four filters for eachof the tubes of input 35S (step 6), and two filters for each of the tubes of incorporated 35S(step 12). For the control salmon DNA and the input 35S samples, two of each set of fourreplicate filters will be subjected to TCA precipitation and washing, while the other twowill be put directly into scintillation vials as “unwashed” samples. For the incorporated35S samples, the two replicate filters will be subjected to TCA precipitation and washing.

19. Mix the tubes of salmon sperm DNA and spot 5 µl of each mixture onto the corre-sponding two or four replicate filters. Allow the spotted samples to air dry or use aheat lamp.

20. Use forceps to place duplicate filters from the control salmon DNA and the input 35Ssamples in individual scintillation vials and set aside as unwashed samples.

21. Place the remaining duplicate filters from the control salmon DNA and the input 35Ssamples, along with the duplicate filters from the incorporated 35S samples, into a250-ml glass or metal beaker with 50 ml of 10% ice-cold TCA.

This will precipitate the nucleic acids and wash away unincorporated rNTPs. Up to 18filters can be washed with 50 ml; for more filters, scale up the volumes proportionally.


15.8.15


22. Swirl the beaker on ice for 10 min, then pour off the TCA.

Continual swirling ensures that filters do not clump together.

CAUTION: The TCA and the methanol wash will contain unincorporated rNTPs andshould be disposed of as hazardous radioactive waste.

23. Repeat steps 21 and 22 twice more.

24. Add 50 ml cold methanol to the filters, swirl on ice for 5 min, and pour off methanol.

25. Use forceps to spread out the washed filters on a sheet of plastic wrap and allow to dry.

A heat lamp can be used to hasten this process.

26. Use forceps to transfer these filters to individual scintillation vials.

27. Add 5 ml scintillation cocktail to the vials of washed and unwashed filters andmeasure counts per minute (cpm) in a liquid scintillation counter.

The washed and unwashed control filters and the washed input 35S filters should containonly background levels of radioactivity (<200 cpm), confirming that unincorporatednucleotides were efficiently removed.

28. Average the cpm from duplicate filters and subtract background counts:

input cpm = (unwashed input 35S filters) – (unwashed control filters)incorporated cpm = (washed incorporated 35S filters) – (washed control filters)

29. Determine the fraction of [α-35S]CTP that was incorporated according to the fol-lowing equation using volumes at time of sampling:

This calculation assumes a uniform product length. The NTP purity term refers to thefraction of 35S present in intact NTP molecules. This amount is typically 0.90 to 0.99 infresh radiochemical preparations. See Commentary for more details.

At least 30% of radiolabel should be incorporated for simple templates and at least 10%for difficult templates with a high degree of secondary structure.

30. Multiply this fraction by the total amount (nmol) of radiolabeled plus unlabeled CTPin the reaction to obtain the amount in nmol of cytidine incorporated into productRNA.

31. Divide by the fraction of cytidine residues in the transcript to calculate the total nmolof ribonucleotides incorporated into product RNA.

32. Multiply this value by 10–9 mol/nmol and by Avogadro’s number (6.022 × 1023

molecules/mol) to obtain the total number of incorporated ribonucleotides.

33. Divide by the transcript length to determine the number of transcripts composing thepurified RNA.

34. Make duplicate serial dilutions of one or more transcripts quantified in this mannerand process them through the DNase treatments and reverse transcription in parallelwith the experimental RNA samples.

By using duplicate standards, one can ensure that the protocols used to prepare cDNAare quantitative and reproducible. Five or six serial 10-fold dilutions are usually suitable.Exact quantities depend upon the anticipated range of expression of the GOI in theexperimental samples.


QuantitativeReverse

Transcription PCR

15.8.16


These standard RNAs must be processed in as similar a manner as possible to that ofthe experimental samples to avoid introducing discrepancies in the efficiency of reversetranscription and PCR amplification of RNA standards versus experimental RNA samples.For example, mock-infected tissue or cell homogenates could be added to RNA standardsused for assays of mRNAs from infectious agents. Unrelated RNA such as yeast or E. colitRNA may be used (at concentrations that mimic the total RNA content of experimentalsamples) in cases where preparations free of the target RNA species are not available.If multiple RNA species will be assayed in single experimental samples, the differentsynthetic RNAs may be combined as a single, serially diluted RNA standard.

35. Employ the resulting duplicate set of standard cDNA samples in the standard curvemethod (see Basic Protocol 1) as the dilution series of standard cDNAs to obtainabsolute quantification of RNA species in unknown samples.

Generate standard curve by plotting Ct against the logarithm of input RNA copy numberfor the RNA standards. Linear regression performed on these points yields an equationfrom which the copy number of RNA in an unknown sample can be calculated from its Ct.

SUPPORTPROTOCOL 2

DESIGN AND VALIDATION OF SYBR GREEN AND TaqManPRIMER/PROBE SETS

Considering that QPCR relies on the quality and the fidelity of the primers and probes thatare used, very strict parameters for their design and subsequent validation are required.A common misconception in performing QPCR assays is that if the primer works fortraditional end-point PCR, it is suitable for QPCR. In some cases this is true. However,the primer set must be tested in a QPCR validation assay before it can be used forRNA expression analysis. In keeping with the high-throughput capacity of QPCR, thethermocycling conditions are kept constant for all assays: 10 min at 95◦C activates thehot-start Taq polymerase, followed by 40 cycles of the two-step 95◦C melting and 60◦Cannealing. An extension step in the thermocycler program is not required, since all ofthe PCR products are 50 to 150 bp, thus making the run last only 1.5 to 2 hr. Theconcentrations of PCR reagents such as Taq DNA polymerase, MgCl2, other salts, anddNTPs remain constant within the same chemistry (i.e., SYBR Green or TaqMan), andthese so-called universal cycling conditions make primer and probe sequences the onlypoint of flexibility in performing the assays.

The design of primer/probe sets requires the availability of reliable sequence informa-tion that may be obtained from databases like NCBI’s GenBank or Ensembl, or fromdata produced by direct sequencing. The assay does not tolerate base mismatches be-tween primer and template, especially in the probe sequence, a feature that allows forthe detection of single-nucleotide polymorphisms (SNPs). Several primer/probe designsoftware packages are available either for purchase or online. Otherwise, the user maydesign the primer sets by directly examining the sequence and choosing primers withthe correct characteristics, as outlined in this protocol. The probe is labeled at the 5′ endwith a fluorescent reporter such as 6-FAM or VIC, and at the 3′ end with a fluorescentor nonfluorescent quencher. The user should consult with the vendor that will synthesizethe probe for the availability of each type of label.

The assay consists of a standard cDNA dilution series from which linear regressioncurves may be plotted. The slope of the resulting curve gives a measure of PCR effi-ciency, where −3.3 ± 0.1 with a coefficient of determination (R2) of 0.99 indicates areaction efficiency of 100%. Part of the initial SYBR Green validation also includes amelting (or dissociation) curve analysis. At the end of the repetitive cycles of the PCR,an additional melt-anneal-melt cycle is performed. The final melt occurs very slowly andthe changes in both temperature and fluorescent signal are monitored over time. Thisdecrease in fluorescence correlates with the dissociation of the double-stranded PCRproduct releasing the bound SYBR Green I fluorophores. The instrument software uses


15.8.17


an algorithm to transform and display the melting curve as the negative first derivativeof the normalized fluorescence versus temperature (Applied Biosystems, 2001a). Thepresence of a single peak in the melting curve is indicative of a single PCR product, andoccurs at the melting temperature of the product. Multiple peaks in this plot indicate thatnonspecific products or primer dimers have been formed. Formation of a single productcan be confirmed by running the PCR products on a 2% agarose or 10% polyacrylamidegel following the QPCR run. Occasionally, when two products are observed, the secondproduct may have been formed during the plateau phase, which would not affect quan-tification. To confirm whether this has occurred, the QPCR run could be repeated andstopped during the exponential phase, and the reaction products run on an agarose gel.However, this is not feasible in practice because of the high-throughput nature of theassay. The best course of action when multiple products are observed in the dissociationcurve is to redesign and validate a new primer set. Only primer sets that give a singlepeak in this curve should be used for experimental assays. Once a SYBR Green–basedprimer set has passed validation testing, the corresponding TaqMan probe is ordered andvalidated for PCR efficiency only. In rare cases, the SYBR Green assay conditions (e.g.,primer concentration, Mg2+ concentration) will not be appropriate for TaqMan assays.This is observed as a decline in PCR efficiency. In this case, new primers may be designedto flank the probe sequence.

Additional Materials (also see Basic Protocol 1)

Primer/probe design software (Primer Express, Applied Biosystems)

Design primers1. Retrieve the RNA sequence information from the appropriate source (e.g., Genbank

or Ensembl).

2. Determine the locations of exon boundaries by aligning the mRNA sequencewith its gene or by using NCBI’s Entrez Gene Evidence Viewer (http://www.ncbi.nlm.nih.gov) or Ensembl’s Genome Browser (http://www.ensembl.org).

Some genes do not have introns, so this step may not be applicable.

3. Copy the sequence into the design software.

Several design programs are available both commercially as stand-alone applicationsand as Web-based applications. Alternatively, primers and probes may be designed “byhand.” Omit this step if designing by hand.

4. If using software other than Primer Express, use the following parameters:

a. QPCR primers: Should have 40% to 60% GC content and melting temperaturesaround 60◦C. Should not contain runs of the same nucleotide, repetitive sequences,or more than two G’s and/or C’s on the 3′ end (also called GC clamp).

b. PCR product (amplicon): Should be 50 to 150 bases in length with an approximatemelting temperature between 85◦ and 95◦C.

c. TaqMan probe (anneals to sequence between primers): Should have the sameproperties as the primers, except that the melting temperature should be around70◦C, and the sequence should not contain G’s within a few bases of the 5′end because of increased reporter quenching. In addition, the sequence must havemore C’s than G’s, which can be accomplished by using the complementary strandsequence for the probe (Applied Biosystems, 2002b).

Primer Express contains templates into which these parameters have been preloaded.


QuantitativeReverse

Transcription PCR

15.8.18


5. Choose a primer/probe set for which the primers anneal in different exons, or whichhave less than 5-bp overhangs into the adjacent exon on the 3′ end of the primer.

This step is necessary to avoid amplification of contaminating genomic DNA. Althoughthe RNA is DNase-treated prior to reverse transcription, complete removal of genomicDNA is never achieved.

For intron-less transcripts or other primer sets that bind sequence within a single intron,the –RT controls are essential for each sample, to ensure that genomic DNA is not beingamplified.

6. Perform a BLAST (or equivalent) search of both primers of the set together to verifythat they will fully anneal to the correct sequence and only that sequence.

7. If the TaqMan probe will be used, run BLAST (UNIT 19.3) on the probe sequence toassess whether it binds to the correct sequence with 100% identity.

8. Order a small-scale synthesis of the primers from a suitable vendor. Standard de-salting of the primers is sufficient, and no additional purification (e.g., HPLC) isrequired.

Once a primer set has been validated, large-scale synthesis may be more cost effective,especially for frequently used primers like the reference genes.

9. Validate the primer set according to the remaining steps of this protocol. If performingTaqMan-based assays, validate the primer set before ordering the probe. Once theprimer set is validated, order the dual-labeled probe from an appropriate vendor, andvalidate according to the following steps.

Figure 15.8.3 Typical plate setup for primer and probe validation assays. The plate arrangement shows a standardcDNA template dilution series that is being used to test 15 primer sets along with an endogenous control primer setthat has already been validated. The same format is used when testing new probe sets. If different standard cDNAsare used to test primers/probes on the same plate (i.e., standards derived from different tissues), the validatedendogenous reference primers/probes must also be run for that standard.


15.8.19


Table 15.8.3 Master Mixes for Primer/Probe Validation Assays


Volume perwell


Master mix (for 8standardsb + 1 extra)

SYBR Green assay

2× SYBR Green mix 1× 5 µl 20 µl 180 µl


150 nM eachprimer

1.2 µl 4.8 µl 43.2 µl

Template cDNA 0.016-50 nga 1.25 µl 5 µl —

H2O N/A to 10 µl to 40 µl 91.8 µl


TaqMan assay

2× TaqMan mix 1× 5 µl 20 µl 180 µl


300 nM eachprimer

2.4 µl 9.6 µl 86.4 µl

5 µM TaqMan probe 250 nM 0.5 µl 2.0 µl 18 µl

Template cDNA 0.016-50 nga 1.25 µl 5 µl —

H2O N/A to 10 µl to 40 µl 30.6 µl


aRecommended template dilution series. The user may modify the range of cDNA concentrations based on the experimental system. Note thatthe cDNA template quantity is based upon the amount of total RNA input into the reverse transcription reaction.bFor this assay, the NTC and –RT controls are included as part of the standard sample set.

Validate primer set10. Set up assay as described in Basic Protocol 1 (steps 1 to 9), except refer to Figure

15.8.3 for an example of a typical plate setup and to Table 15.8.3 for master mixcomponents.

11. Test the new primer set using SYBR Green chemistry. If valid, test the TaqMan-basedchemistry.

12. Following the instrument run, first check the dissociation curve. If more than onepeak is present, the primer set is invalid and no other parameters are checked.

It may be possible to eliminate the nonspecific product(s) detected in the dissociationcurve analysis or gel electrophoresis by addition of enhancers such as betaine (UNIT 15.1).This approach may be warranted if the sequence constraints for a given GOI are limiting.

13. If a single peak is found in the dissociation curve, assess the PCR efficiency bycalculating the slope of the linear regression curve as follows.

a. Plot Ct (or crossing point, CP) versus log ng of the standard cDNA input for eachconcentration of standard as an xy scatter plot.

This may be performed directly in the instrument software on the Applied Biosystemsinstrument, or can be done in Excel (or equivalent).

b. Apply a best-fit curve and display the corresponding linear regression formula.

If the slope of the curve is −3.3 ± 0.1 with R2 = 0.99, the primer set amplifies at100% efficiency, and the set is considered valid.


QuantitativeReverse

Transcription PCR

15.8.20


Efficiency is dependent upon several factors including pipet calibration, primerquality and dilution, and even variability in the instrument run. The slope, therefore,may not be exactly −3.3 ± 0.1. In this case, the slope of the test primer set mustmatch that of a previously validated endogenous reference gene run for the samestandard cDNA within ±0.1. For example, if the test primer gives a slope of −3.6 andthe endogenous reference for the standard cDNA used to test the set gives a slope of−3.5, then the test set is valid. However, if the endogenous reference primer set givesa slope of −3.3 and the test primer has a slope of −3.6, the test set is invalid.

14. Repeat the linear regression analysis for the TaqMan probe/primer set to assess PCRefficiency.

Melting curve analysis cannot be performed for TaqMan-based assays, since cleavage ofthe probe releases the reporter that continuously fluoresces.

REAGENTS AND SOLUTIONSUse molecular-biology-grade (nucleic acid– and nuclease-free) or sterile-filtered double-deionized water in all recipes and protocol steps. For common stock solutions, see APPENDIX 2;for suppliers, see APPENDIX 4.

Primer mixes, 1.25 µM each forward and reverse primerMix small aliquots of 2.5 µM forward and reverse primer stocks (see recipe) inequal volumes (1:1). Store up to 1 to 2 months at 4◦C in screw-capped tubes toprevent evaporation. Do not freeze, as repeated freeze-thaw cycles will degrade theprimers.

Primer stocks, 100 and 2.5 µMPurchase lyophilized oligonucleotides from any commercial source (synthesizedat a 25-nM scale; standard desalting is sufficient, no additional purification isneeded). Briefly centrifuge the tubes of powdered oligonucleotide to get contentsto the bottom. Resuspend in molecular-biology-grade water to 100 µM. Beforepreparing primer mixes, dilute each stock (forward and reverse) to 2.5 µM. Store at–20◦C. When stored properly and subjected to minimal freeze-thaw cycles, primerstocks can last >2 years.

Primer pairs produced at a 25-nM scale will yield enough reagent to test ∼4000 samplesusing SYBR Green, or ∼2000 samples using TaqMan assays.

Sterile-filtered double-deionized water may be used instead of purchased molecular-gradewater. Do not use DEPC-treated water, because the slightly acidic pH may promote primerdegradation. Tris buffer may be used instead of water, but should not contain EDTA, whichacts as a Mg2+-chelating agent and can inhibit the PCR.

Standard cDNAs, dilution seriesDNase-treat (UNIT 3.12) and reverse transcribe (UNIT 3.7) a suitable RNA using 0.08µg/µl random hexamer primers such that the final concentration of the standard is40 ng/µl (based on RNA quantity; see UNIT 15.5). Include both a no-template control(NTC) for which water is used instead of RNA, and a no-reverse-transcriptasecontrol (−RT). Following the reverse transcription, make a 5-fold dilution series ofthe 40 ng/µl standard to obtain working concentration standards of 40, 8, 1.6, 0.32,0.064, and 0.0128 ng/µl. Store up to 1 month at 4◦C or >2 years at –20◦C (withminimal freeze-thaw cycles).

By using 1.25 µl/well of each of the standards in the final reaction plate, the resultingamount of starting template will be 50, 10, 2, 0.4, 0.08, and 0.016 ng. See Commentary formore information and suggestions about control and standard RNA samples.

Alternatively, if absolute RNA standards are being prepared via Support Protocol 1, theseshould be serially diluted first and then, in parallel with experimental RNA samples, DNase-treated and reverse transcribed (see Support Protocol 1, steps 34 to 35).


15.8.21


Table 15.8.4 Preparation of 2× SYBR Green Mix

Component Volume (µl)

25 mM MgCl2 48 288 600 720 1200

10× Gold PCR buffer 40 240 500 600 1000

10× dNTP mix 40 240 500 600 1000

DMSO 40 240 500 600 1000

1:1000 SYBR Green I 20 120 250 300 500

50× ROX 8 48 100 120 200

AmpliTaq Gold polymerase (5 U/µl) 2 12 25 30 50

H2O 2 12 25 30 50

Total volume in 2× buffer 200 1200 2500 3000 5000

SYBR Green mix, 2×Combine the following components as indicated in Table 15.8.4:25 mM MgCl2, molecular biology grade (store at –20◦C)10× Gold PCR buffer (supplied with PCR enzyme; Applied Biosystems; store at

–20◦C)10× dNTP mix: equal volumes of 2 mM dATP, dTTP, dCTP, and dGTP (store at

–20◦C)Dimethylsulfoxide (DMSO), molecular biology grade (store at room temperature)SYBR Green I dye (Molecular Probes; store at –20◦C), diluted 1:1000 in water50× ROX passive reference dye (Invitrogen; store at –20◦C)AmpliTaq Gold polymerase (5 U/µl; Applied Biosystems; store at –20◦C)Water, molecular biology grade

Prepare fresh and keep on ice prior to use in primer master mixes. Protect dyesand all prepared mixes from prolonged exposure to light by wrapping tubesin foil.

To maintain the high-throughput nature of the assay, all buffer conditions, including theconcentrations of Mg2+, dNTPs, and other additives, are kept constant. It is stronglyrecommended that a preformulated buffer be purchased from a reliable vendor, since theseare stable at room temperature and have been quality-control tested to ensure optimalperformance. The authors have obtained comparable results using mixes from AppliedBiosystems, Bio-Rad, Invitrogen, and Sigma. However, it is important to compare lots, asthere can be lot-to-lot differences in the commercial preparations.

TaqMan probe, 100 and 5 µMDepending on the vendor, the probe may be supplied in a lyophilized form. In thiscase, resuspend to 100 µM in Tris·Cl, pH 8.0 (APPENDIX 2), prepared with molecular-biology-grade water and reagents. Before use, dilute a small amount of 100 µMstock to 5 µM with more Tris·Cl, pH 8.0. Store either concentration at –20◦C. Avoidrepetitive freeze-thaw cycles, and thaw on ice prior to use to preserve the integrityof the probe. Protect the probe from excessive exposure to light (e.g., using amber-colored screw-cap tubes) to prevent photobleaching of the fluorescent dyes andevaporation. See Support Protocol 2 and Commentary for additional considerationsregarding the TaqMan probe.


QuantitativeReverse

Transcription PCR

15.8.22


Trichloroacetic acid (TCA) solution, 10% (w/v)Prepare a 100% (w/v) TCA stock solution by dissolving the entire contents of anewly opened TCA bottle in water (e.g., dissolve the contents of a 500-g bottle ofTCA in sufficient water to yield a final volume of 500 ml). Store up to 1 year at4◦C. Prepare 10% (w/v) TCA by dilution and store up to 3 months at 4◦C.

CAUTION: TCA is extremely caustic. Protect eyes and avoid contact with skin whenpreparing and handling TCA solutions.

COMMENTARY

Background InformationQuantitative PCR is a rapid, robust, and

highly sensitive polymerase chain reactionmethod used to quantify specific nucleic acidtargets. Real-time quantitative PCR is differ-ent from end-point, or in-gel, analysis (seeUNIT 15.7) in several ways. For real-time analy-sis, the increase in fluorescent signal resultingfrom PCR product synthesis is recorded dur-ing the course of the thermocycle. This allowsthe user to specify the point in the assay atwhich to “read” the data. Measurements areobtained from the geometric phase of the am-plification reaction. This is the phase duringwhich all of the components required for thePCR (e.g., dNTPs, primers, polymerase) arein excess, and therefore the deficit of an essen-tial reaction component will not quench theefficiency of product synthesis. Following ge-ometric amplification, the fluorescence curvereaches a plateau (i.e., the saturation point)as the reaction components begin to becomelimited and the kinetics of the reaction becomeunpredictable. At this stage, an increase of onethermocycle no longer correlates with a two-fold change in product (Applied Biosystems,2002a). In contrast, in end-point PCR, quan-tification is often obtained by in-gel densit-ometry measurements at the end (or the sat-uration point) of the reaction where the re-action may be at a plateau, compromisingquantification.

Real-time PCR depends both on a set ofuniversal thermocycling and buffer conditionsand on primer efficiency testing and correc-tion where necessary. As a result, the accu-racy and precision of the resolution (small-est detectable fold-change) is less than 2-fold, whereas resolution for end-point in-gel measurement is limited to about 10-fold(Applied Biosystems, 2002a). PCR assaysin which samples are removed at measuredcycle times and electrophoresed and possi-bly hybridized have better resolving powerthan end-point analysis. However, both in-gelmethods suffer from the lengthy processingsteps, compared to real-time PCR.

The two most widely used fluorescent de-tection methods, or chemistries, for QPCR areSYBR Green, a DNA-intercalating dye, andthe fluorogenic probe. Several types of fluoro-genic probes are currently available: the pop-ular, dual-labeled hydrolysis probe (TaqManprobe) and the hybridization probes known asmolecular beacons and scorpions. Both typesof probes bind the sequence intervening theforward and reverse primer binding sites, andboth rely on fluorescence resonance energytransfer (FRET) to silence the signal from thereporter dye while it is in proximity to thequencher dye. A hydrolysis probe is cleavedby the 5′ nuclease activity of the polymeraseduring the primer-extension phase of the reac-tion, and the reporter is released and becomesfree to fluoresce continuously. Hybridizationprobes rely on a stem-loop structure to keepthe reporter and quencher in proximity. Uponhybridization to the specific sequence, the dis-tance between the reporter and quencher be-comes too large to silence the reporter, andsignal is detected (Tyagi and Kramer, 1996;Whitcombe et al., 1999).

There are advantages and disadvantages toeach type of chemistry (SYBR Green versusTaqMan) in terms of sensitivity and specificity.SYBR Green I binds any double-strandedDNA and does not depend on a probe-cleavageevent. Therefore, SYBR Green produces ear-lier Ct values, resulting in an apparent en-hanced sensitivity (Whittwer et al., 1997; Mor-rison et al., 1998). TaqMan probes, on the otherhand, supply another layer of sequence speci-ficity in addition to the forward and reverseprimers.

There are two methods of quantificationthat may be performed using real-time PCR:absolute and relative. Absolute quantificationmeasures the copy number of a specific nu-cleic acid target in a sample. Relative quantifi-cation measures the difference in copy num-ber between two samples that have each beennormalized to an endogenous reference. Bothcan be used to compare the effects of differenttreatments on a particular RNA species or to


15.8.23


compare the levels of multiple RNA species ina single sample. The absolute method requiresstandards in which the copy number of theparticular target has been carefully and accu-rately measured. From this standard sample, adilution series is made and assayed for the tar-get at the same time as the unknown samples.From the values obtained from linear regres-sion analysis of the standard dilution series, theGOI copy number values may be interpolated.This will allow the user to assess the sensitiv-ity (i.e., the lowest detectable copy number) ofthe GOI primer set. As an example, this typeof analysis is used in both clinical and foodscience for the assessment of pathogen loadand gene copy number (Pfaffl, 2004).

By contrast, relative quantification does notrely on the knowledge of a given transcriptcopy number in a standard sample. Instead,the changes in gene expression or the lev-els of a specified transcript may be measuredand described as an arbitrary unit relative tosome control sample (��Ct or standard curvemethod), or to the level of some other controltranscript in the same sample (standard curveor efficiency-corrected �Ct method). For ex-ample, relative quantification allows for themeasurement of the fold-change in expressionfor gene A in a treated sample versus an un-treated sample, or for the assessment of thelevel of gene A in a sample relative to somehousekeeping gene in the same sample.

Critical Parameters andTroubleshooting

Primer/probe design and validationFor RNA analysis, it is important to differ-

entiate message from genomic DNA. In this re-spect, the use of amplicons that span an exonjunction allows this requirement to be met.This is achieved by designing primer sets withthe forward and reverse primers sitting in dif-ferent exons. However, when the interveningintron is small, genomic DNA may still be am-plified. This can be avoided by allowing a fewbase pairs on the 3′ end of either primer tooverhang onto the neighboring exon.

To distinguish knockout or mutant samplesusing QPCR, primers/probes should be de-signed in the knocked-out or mutated regionof the transcript. In this way, no amplificationof the transcript will occur for the knockout,since the region recognized by the primers hasbeen deleted or altered. Primers for transcriptsthat are rapidly degraded should be placed nearthe 5′ end of the RNA sequence, as degrada-tion by ribonucleases generally occurs 3′ to

5′ (Brown, 2002). However, RNA transcriptsthat have undergone linear RNA amplification(for instance, RNA isolated from laser capturemicrodissected cells) are around 200 to 1000bases in length and represent the 3′ ends ofthe transcripts. In this case, primers must bedesigned in a region near the poly(A)+ tail.

It is absolutely mandatory to validate bothprimers and probes on the same instrumentused to perform the experimental assay, i.e.,when the instrument used to validate theprimer sets is a different brand than the onethat will be used for the experimental assays(e.g., Roche iCycler versus ABI 7900HT).Published primer/probe sets for which the PCRproduct is ≤150 base pairs and the annealingtemperature is around 60◦C, which have beenvalidated properly, should be transferable toany system, but this must be empirically de-termined to ensure the reliability of the data.

In rare cases, the PCR efficiency of a vali-dated primer set changes when switched fromSYBR Green to TaqMan-based chemistry.This is due to the difference in primer concen-tration and/or MgCl2 concentrations betweenthe two buffers, and may often be overcomeby redesigning the primers to recognize a se-quence around the already synthesized probe.

Choice of chemistryThe selection between SYBR Green I and

TaqMan-based assays depends upon the RNAsequence. If the RNA of interest has no poly-morphisms or other variations in the region towhich the primers bind, and the primers havebeen correctly designed and validated, thenSYBR Green–based assays are adequate forgene-expression analyses. It is up to the userto decide if the addition of the often-costlyTaqMan probe is worth the additional speci-ficity it confers. In the case of polymorphismsor variants, differentiation between differentRNA species may require the specificity ofthe probe, which can discriminate a single basedifference.

Template qualityThe quality of the cDNA template depends

upon the integrity of the RNA. QPCR will tol-erate some degradation of the RNA when ran-dom hexamers (or other -mers) are used toprime the reverse transcription reaction. How-ever, it is not good laboratory practice to usedegraded RNA, and the cause of the degrada-tion should be addressed. Transcripts decay atdifferent rates and have variable stability, sopartial degradation of a sample at any pointcould lead to complete absence of detection of


QuantitativeReverse

Transcription PCR

15.8.24


the desired target RNA in a subsequent QPCRassay, with little to no change in the house-keeping gene used as a control.

Total RNA is used for QPCR to reducethe number of steps and potential sources ofdegradation during sample preparation. Puri-fied messenger, or poly(A)+, RNA can alsobe used. However, this subtractive purificationcould lead to the loss of transcripts that donot have a poly(A)+ tail, or to the preferen-tial enrichment of RNAs that have internal Atracts. The added processing reduces the re-covery of material for subsequent use, and cancause degradation. If poly(A)+ RNA is used,50- to 100-fold less (∼100 pg) sample tem-plate is required in the reaction mixture be-cause mRNA represents 3% of total cellularRNA (Alberts et al., 1994).

Reverse transcription primersReverse transcription (RT) of RNA into

complementary DNA (cDNA; refer to UNIT

15.5) may be performed using several differenttypes of oligonucleotide primers. For QPCR,the preferred primer is a random hexamer, non-amer, or dodecamer oligo with 6, 9, or 12-base stretches of random sequences, respec-tively. Random primers have a much higherprobability of efficiently amplifying all RNAtranscripts, due to their indiscriminate nature(ABI, pers. comm.). This property also enablesRNA secondary structure to be overcome,since the priming occurs in random placesalong the length of the transcript. mRNA-specific priming by oligo(dT)s at the poly(A)+

tail and any internal poly(A) tract is anothermethod of RT priming. This method will allowonly polyadenylated RNAs to be convertedto cDNA, thus limiting amplification of someGOIs or partially degraded samples. A mixtureof a random oligo with an oligo(dT) primermay enhance detection of rare messages whilestill allowing for the detection of transcriptsthat lack polyadenylation. However, the use ofthis procedure may skew the measurement ofrelative RNA abundance towards intact, full-length mRNA over incomplete or rapidly de-graded messages. Users should decide and testthese parameters in their particular experimen-tal systems. The third method of reverse tran-scription priming is the gene-specific reverseprimer. The reverse PCR primer is used tospecifically target the GOI for conversion tocDNA. This is often performed in the sameQPCR plate as the PCR by adding RT en-zyme to the PCR mix and adding an incuba-tion step prior to the first step in the PCR cy-cling program (known as one-step RT-PCR).

While this may enhance detection of a specificRNA target, RNA secondary structure maynot be overcome, depending on the primingsite, and RT efficiency must be considered inaddition to PCR efficiency.

An important assumption that is made whenperforming RT-QPCR is that RT efficiencyis similar for the GOI between samples andfor different GOIs. However, different tis-sues/sample types may contain variable levelsof RT-inhibiting or -enhancing factors (Pfaffl,2004). To control for these variables, it is rec-ommended that all samples to be compared beprepared under the same conditions and at thesame time (i.e., with regard to RNA extractionmethod and RT reaction).

Endogenous reference geneNormalization of sample loading is essen-

tial in any quantitative comparative analysisto ensure that the measured differences be-tween samples is not attributable to dispro-portionate amounts of starting material. Forgene-expression assays, the normalizer mustbe an endogenous gene that is expressed atequal levels in all tissue or cell types and treat-ment conditions under study. Traditionally,one of the so-called housekeeping genes (e.g.,GAPDH, cyclophilin, β-actin, HPRT, U36B4,18S rRNA) is selected to serve this function.The choice is a point of controversy, sincethere are examples of fluctuations for most ofthe abovementioned genes under various treat-ment or physiological conditions (Schmittgenand Zakrajsek, 2000; Suzuki et al., 2000; Guoet al., 2001; Vandesompele et al., 2002; Dhedaet al., 2004). The user should run a small pi-lot experiment to determine which endogenousreference is appropriate for the particular stud-ies. A good method for doing this is to performa ��Ct assay using several potential normal-izer RNAs and a GOI that is not expected toshow any fold-changes within a small set ofsamples. The GOI is then normalized to eachof the reference RNAs individually for all ofthe samples. Finally, the housekeeping genefor which there is no detectable fold-changeof the GOI between the test samples is cho-sen for use in experimental assays (AppliedBiosystems, 2001b; Guo et al., 2001; RocheApplied Science, 2002).

Controls and relative RNA standardsControls for the assay that are made along-

side the cDNA standards and unknowns in-clude a no-template control (NTC), made bysubstituting water for RNA, and a no-reverse-transcriptase (–RT) control, made by omitting


15.8.25


the reverse transcriptase. For primers that spanan exon junction, a –RT control is not neededfor every sample if, during the validation pro-cess, this control shows no amplification prod-uct. If the system under study is the result ofintroduction of an expression construct intocells, or if the GOI does not have introns orhas a known processed pseudogene, then –RTcontrols should be made and assayed for ev-ery sample. An NTC should be run for everyprimer set of an assay to facilitate the detectionof contaminants that contribute to fluorescentsignal.

If amplification of the NTC occurs, primer-dimer or other nonspecific PCR products mayhave been formed, or contamination of areagent or degradation of the primer mix mayhave taken place. If amplification occurs inthe –RT control, this indicates the presenceof genomic DNA if the primer/probe set doesnot span an exon junction, or the presence ofprimer-dimer or nonspecific PCR products, acontaminant, or degradation of the primer mix.Although all amplification products (includ-ing those that are nonspecific) contribute tothe fluorescence signal, this fluorescence con-tribution can be considered negligible if the Ctvalues of the NTC and –RT are ≥7 cycles dif-ferent from the experimental samples (AppliedBiosystems, pers. comm.).

If the unknowns are RNA samples, the stan-dard of choice is an RNA that is reverse tran-scribed in the same manner as the unknowns.A suitable standard RNA is one in which theexpression of the GOI is at a moderate tohigh level and that has a similar composi-tion to the unknown samples. The use of aplasmid or linear DNA standard is not ad-vised for measurement of an endogenous tis-sue transcript, because these types of nucleicacids have different background compositionsthan RNA, are extracted differently, and arenot reverse transcribed, and therefore maynot have the same RT or PCR amplificationefficiency as the experimental samples (Ap-plied Biosystems, 2003; Pfaffl, 2004). Severalvendors (e.g., Ambion, Clontech, Stratagene)supply total RNA preparations of many celland tissue types collected from many differentspecies. Both Stratagene and Clontech maketotal RNA pools termed “universal referenceRNA,” composed of mixtures of either vari-ous cell lines or whole tissues, respectively.These RNA pools represent ≥90% gene cov-erage on microarrays (Novoradovskaya et al.,2000; Clontech, 2002), and are very useful asstandards both for validation of primer/probesets and for large-scale multigene studies.

Absolute RNA standardsThe absolute quantification of RNA syn-

thetic standards by radiolabel incorporationpermits accurate determination of transcriptamounts. Liquid scintillation counting of thefilters typically yields incorporation valueswith errors of less than ±5 %. However, er-rors of up to 10% to 15% will not significantlyaffect quantification in most applications. Theuse of low-energy 35S (compared to 32P) atvery low specific-activity levels minimizes po-tential radiolytic degradation of the syntheticRNA over time and minimizes personal radia-tion exposure. Very small amounts of standardRNA are typically needed for an RT-PCR as-say (less than 108 copies), such that the amountof 35S in each RT-PCR reaction will be ex-tremely small.

One consideration is the purity of the radio-labeled ribonucleotide. Fresh lots are generallyguaranteed by the manufacturer to be 90% to99% intact NTP. The presence of other con-taminating radiolabeled material is taken intoaccount when calculating total input 35S. Stor-age time and multiple freezing and thawingcycles will contribute to decomposition of theradionucleotide. If the integrity of the reagentis in doubt, consult the manufacturer, assessthe fraction of intact NTP by thin-layer chro-matography, or order a fresh lot.

While RNA can be quantified by opti-cal density, care must be taken to eliminateunincorporated nucleotides (which will alsoabsorb light at 260 nm) and take into ac-count RNA secondary structure (which can re-duce absorbance). A protocol for quantifyingnon-radioactive synthetic RNAs that accountsfor secondary structure by measuring the ab-sorbance of hydrolyzed RNA is described inIyer and Struhl (1996). This alternative may besuitable for researchers who prefer not to han-dle radioactive compounds. RNA can also bequantified by comparison with mass standardson a gel; however, this method may be inaccu-rate due to the differential binding of ethidiumbromide to single- and double-stranded RNAregions.

The most important concern is to avoidRNase contamination of equipment andreagents. If RNA yield from transcription islow or undetectable, RNA degradation is themost likely culprit. Gel electrophoresis of theRNA samples can distinguish transcriptionfailure (no products in the pre-DNase sample)versus loss of RNA during subsequent manip-ulations (products in the pre-DNase samplebut not in the purified sample). More frequentsamples can be taken for gel analysis to help


QuantitativeReverse

Transcription PCR

15.8.26


pinpoint the troublesome step or reagents. Aribonuclease inhibitor such as SUPERase·In(Ambion) may be included in the transcrip-tion reaction and added to the purified RNAto preserve RNA integrity; however, this willmerely mask RNase contamination, not elim-inate it. Other possible solutions for poor in-corporation or yield could be old reagents. Payparticular attention to the freshness of the [α-35S]NTP, DTT, rNTPs, and polymerase.

A single RNA product species from thetranscription is critical for the production ofaccurate RNA standards. Multiple productspecies of indeterminate content can cause in-accuracies in quantification and can also leadto extraneous PCR products. Product RNAspecies longer than expected indicate incom-plete cleavage of the parent plasmid. Either in-creasing the efficiency of the plasmid digestionor purifying the linearized plasmid from anagarose gel will prevent this problem. Speciessmaller than expected could indicate that theRNA polymerases had difficulty transcribingfull-length RNA. Use of a truncated templatewhich eliminates problematic sequences maybe required. Very minor contaminants may notaffect quantification by more than a few per-cent. The researcher can try to estimate theamount of contaminating bands from the geland account for these in the calculations. How-ever, since the sequence of such contaminantsis unknown, it is difficult to assess the effectduring PCR amplification. Thus, the authorsrecommend striving to attain a single productspecies.

Additional considerations for constructingthe transcription plasmid that will be used toprovide the template for run-off transcriptionare as follows:

1. The cloned sequence from the gene ofinterest must include the sequence primed forreverse transcription and the PCR target se-quence. Ensure that the clone does not includesequences, such as introns, that are not partof the RNA species. Full-length cDNA clonesyield the most authentic synthetic RNAs. Insome cases, truncated cDNAs may permitmore efficient and consistent transcription of asingle-length RNA product; however, obtain-ing a single-length RNA product species iscritical.

2. If oligo-dT is used for reverse tran-scription, then the synthetic RNA will needa poly(A)+ tail. In this case, construct a tran-scription plasmid using a vector such as pSP64Poly(A)+ (Promega), which contains a run ofdA:dT residues at the 3′ end of the multiple

cloning site, allowing for the transcription ofa synthetic poly(A)+ tail.

3. Ensure that the orientation of the clonedcDNA will produce sense transcripts.

4. Engineer the plasmid such that cleavageby a restriction enzyme generates a linear pieceof DNA that contains the phage polymerasepromoter and the entire sequence of the desiredsynthetic RNA. This is ideally accomplishedusing an enzyme which linearizes the plasmidby cutting at only one site, at the end of the de-sired transcript sequence. Avoid including su-perfluous vector sequences in the transcribedregion, since these will not be present in RNAfrom the experimental samples.

Avoid using restriction enzymes that cre-ate 3′ overhangs, because there is evidencethat they cause RNA complementary to the in-tended transcripts to be generated (Schenbornand Mierendorf, 1985).

As an alternative to a plasmid, PCR prod-ucts can also be designed for use as tran-scription templates by incorporating the poly-merase promoter sequence into one of theprimers (Mullis and Faloona, 1987).

Anticipated ResultsThe setup of the ��Ct assay allows 64

samples (one of which should be the NTC) tobe assayed for an endogenous reference geneand one gene of interest in triplicate; the stan-dard curve and efficiency-corrected �Ct as-says will accommodate an endogenous refer-ence gene and one gene of interest in triplicatefor 56 unknowns, six standards, and two con-trol samples. Assays performed on the ABI7900HT are completed in 1.5 and 2 hr for Taq-Man and SYBR Green-based assays, respec-tively. The results yield data that are highlyreproducible and correlate well with tradi-tional northern blotting and RNase-protectionassays. Support Protocol 1 will yield 10 to100 µg of single-length transcripts, equivalentto ∼1012 to 1013 RNA molecules, with a spe-cific activity of ∼1 × 107 cpm/µg.

Time ConsiderationsIn preparation for performing a QPCR as-

say, RNA and subsequent cDNA preparationmay be carried out in advance. Prior to anactual experimental assay, primers and probesmust be designed and validated. In most cases,design and validation may take several days toa few weeks. This includes the time to design,order, and synthesize the primers (∼2 days,depending on the vendor), test the primers (afew hours), and, if desired, synthesize (∼7 to


15.8.27


14 days, depending on the vendor) and test (afew hours) the corresponding probes.

Once the required primer/probe sets are val-idated, the experimental assays are performed.A single-plate assay may take 0.5 to 2 hr to pre-pare and 2 hr to run. The preliminary raw-dataanalyses on the instrument software may takeless than half an hour. The time required for thefollowing final analyses will depend upon theuser’s familiarity with both the mathematicaland software applications. The typical work-flow for an experimental assay is as follows.On the first day, prepare total RNA and de-termine concentration by UV or fluorescencespectroscopy (requiring 1 to 4 hr, dependingon number of samples and method of prepara-tion). Next, DNase-treat and reverse transcribeRNA to cDNA (requiring ∼3 hr for setup andincubations). On the second day, prepare mas-ter mixes for the assay (∼30 min to 1 hr). Next,prepare QPCR plate(s) (∼30 min to 1 hr perplate), run plate on instrument (1.5 to 2 hr),and collect and analyze data (1 to 3 hr).

For preparation of RNA standards, con-struction of the transcription plasmid can takeanywhere from several days to several weeks;one additional day is needed to prepare thetemplate for transcription. RNA synthesis, gelelectrophoresis, and scintillation counting canbe completed the following day. Followingovernight gel drying, the autoradiography mayrequire several hours to several days for suf-ficient exposure. During the incubation pe-riods for transcription and DNase treatment,time becomes available to prepare the gel, fil-ters and scintillation vials. Filters can be spot-ted and washed while the gel is running. Ifnecessary, the samples for gel electrophoresis(dissolved in loading buffer) and/or the sam-ples for scintillation counting may be stored at−20◦C overnight and analyzed the followingday.

AcknowledgmentThis work was supported by the Howard

Hughes Medical Institute and by grantsfrom the Robert A. Welch Foundation (I-1275) and the National Institutes of Health(U19DK62434).

Literature CitedAlberts, B., Bray, D., Lewis, J., Raff, M., Roberts,

K., and Watson, J.D. 1994. The cell nucleus. InMolecular Biology of the Cell, 3rd ed., p. 370.Garland Publishing, New York.

Applied Biosytems. 2001a. ABI Prism 7900HTUser Manual, rev. 4. Applied Biosystems, FosterCity, Calif.

Applied Biosystems. 2001b. TaqMan humanendogenous control plate: Protocol: Revision C.http://docs.appliedbiosystems.com/pebiodocs/04308134.pdf. Applied Biosystems, FosterCity, Calif.

Applied Biosystems. 2002a. Real-time PCR vs. tra-ditional PCR. https://www.appliedbiosystems.com/support/tutorials/pdf/rtpcr vs tradpcr.pdf.Applied Biosystems, Foster City, Calif.

Applied Biosystems. 2002b. Designing Taq-Man MGB probe and primer sets for geneexpression using Primer Express software.http://www.appliedbiosystems.com/support/tutorials/pdf/taqman mgb primersprobes forgene expression.pdf. Applied Biosystems,

Foster City, Calif.

Applied Biosystems. 2003. Creating standardcurves with genomic DNA or plasmidDNA templates for use in quantitative PCR.https://www.appliedbiosystems.com/support/tutorials/pdf/quant pcr.pdf. Applied Biosys-tems, Foster City, Calif.

Brown, T.A. 2002. How genomes function. InGenomes, 2nd ed. (S. Carlson, ed.) section 10.4.John Wiley & Sons, Hoboken, N.J.

Clontech. 2002. Control RNA for microar-ray experiments. Clontechniques XVII:6.http://www.clontech.com/clontech/archive/APR02UPD/pdf/ControlRNA.pdf. Clontech,Palo Alto, Calif.

Dheda, K., Huggett, J.F., Bustin, S.A., Johnson,M.A., Rook, G., and Zumla, A. 2004. Vali-dation of housekeeping genes for normalizingRNA expression in real-time PCR. BioTech-niques 37:112-119.

Guo, D., Henriksson, R., and Hedman, H. 2001. TheiCycler iQ detection system for evaluating ref-erence gene expression in normal human tissue,rev. A. Amplification 2804. http://www.bio-rad.com/LifeScience/pdf/Bulletin 2804.pdf.Bio-Rad, Hercules, Calif.

Iyer, V. and Struhl, K. 1996. Absolute mRNA lev-els and transcriptional initiation rates in Saccha-romyces cerevisiae. Proc. Nat. Acad. Sci. U.S.A.93:5208-5212.

Kramer, M.F. and Coen, D.M. 1995. Quantificationof transcripts from the ICP4 and thymidine ki-nase genes in mouse ganglia latently infectedwith herpes simplex virus. J. Virol. 69:1389-1399.

Morrison, T.B., Weis, J.J., and Wittwer, C.T. 1998.Quantification of low-copy transcripts by con-tinuous SYBR Green I monitoring during am-plification. BioTechniques 24:954-962.

Mullis, K.B. and Faloona, F.A. 1987. Specificsynthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol.155:335-350.

Novoradovskaya, N., Payette, T., Novoradovsky,A., Braman, J., Chin, N., Pergamenschikov,A., Fero, M., and Botstein, D. 2000.Pooled, high-quality reference RNA forhuman microarrays. Strategies 13:121-122.http://www.stratagene.com/news/newsletter.aspx?iid=6. Strategene, La Jolla, Calif.


QuantitativeReverse

Transcription PCR

15.8.28


Pfaffl, M.W. 2004. Quantification strategies in real-time PCR. In IUL Biotechnology Series, No. 5:A-Z of Quantitative PCR (S.A. Bustin, ed.) pp.87-120. International University Line, La Jolla,Calif.

Roche Applied Science. 2002. Selection of house-keeping genes. Technical Note No. LC 15/2002.http://www.roche-applied-science.com/lightcycler-online/lc support/pdfs/lc 15.pdf.Roche Applied Science, Indianapolis, Ind.

Schenborn, E.T. and Mierendorf, R.C. Jr. 1985. Anovel transcription property of SP6 and T7 RNApolymerases: Dependence on template struc-ture. Nucl. Acid. Res. 13:6223-6236.

Schmittgen, T.D. and Zakrajsek, B.A. 2000. Effectof experimental treatment on housekeeping geneexpression: Validation by real-time, quantitativeRT-PCR. J. Biochem. Biophys. Methods 46:69-81.

Shoemaker, J.P., Garland, C.W., and Steinfeld, J.I.1974. Experiments in Physical Chemistry, pp.34-39. McGraw-Hill, New York.

Suzuki, T., Higgins, P.J., and Crawford, D.R.2000. Control selection for RNA quantitation.BioTechniques 29:332-337.

Tyagi, S. and Kramer, F.R. 1996. Molecular bea-cons: Probes that fluoresce upon hybridization.Nat. Biotechnol. 14:303-308.

Vandesompele, J., De Preter, K., Pattyn, F., Poppe,B., Van Roy, N., De Psepe, A., and Speleman, F.2002. Accurate normalization of real-time quan-titative RT-PCR data by geometric averaging ofmultiple internal control genes. Genome Biol.3:1-12.

Whitcombe, D., Theaker, J., Guy, S.P., Brown, T.,and Little, S. 1999. Detection of PCR productsusing self-probing amplicons and fluorescence.Nat. Biotechnol. 17:804-807.

Whittwer, C.T., Herrmann, M.G., Moss, A.A., andRasmussen, R.P. 1997. Continuous fluorescencemonitoring of rapid cycle DNA amplification.BioTechniques 22:130-138.

Key ReferencesAmbion. 2001. The top 10 most common quantita-

tive RT-PCR pitfalls. Technotes Newsletter 8:8.Ambion, Houston, Tex.

A short, but useful checklist of critical considera-tions for performing any type of reverse transcrip-tion PCR.

Applied Biosystems. 1997. Relative quantitation ofgene expression: ABI PRISM 7700 SequenceDetection System: User Bulletin #2: Rev B. Ap-plied Biosystems, Foster City, Calif.

This bulletin outlines both the standard curve and��Ct methods and shows, by comparing data ob-tained using both calculations, that the resultingvalues are very similar regardless of the assay.

Applied Biosytems. 2001a. See above.

This instrument manual contains explanationsabout the transformation of fluorescence signal intoCt data in addition to outlining the proper methodof baseline and threshold settings for ABI machines.

Users should consult their specific instrument man-uals, since each type of instrumentation will requireknowledge of slightly different terminology and pa-rameters.

Livak, K.J. and Schmittgen, T.D. 2001. Analysisof relative gene expression data using real-timequantitative PCR and the 2−��Ct method. Meth-ods 25:402-408.

This article presents a detailed review of the deriva-tion of the mathematical applications described inthis unit.

Internet Resourceshttp://www.ncbi.nlm.nih.gov

NCBI Web site.

http://www.ensembl.org

Ensembl Web site.

http://www.gene-quantification.info/

The Gene Quantification Web site contains a hostof information concerning QPCR.

http://pga.mgh.harvard.edu/primerbank/index.html

The Primer Bank database, hosted by Harvard Uni-versity, contains user-submitted primer sequencesfor several mouse and human genes.

http://web.ncifcrf.gov/rtp/gel/primerdb/

The Quantitative PCR Primer Database (QPPD),maintained by the National Cancer Institute, con-tains primer and probe sequences for mouse and hu-man genes collected from articles cited in PubMed.

http://www.ambion.com/techlib/index.html

Contains numerous, detailed articles and technicalbulletins regarding transcription and general RNAhandling issues.

http://www.promega.com/techserv/

Contains technical manuals with detailed protocoltips and troubleshooting for transcription applica-tions

Other commercial vendor-sponsored technical sup-port Web sites are also a very good resource for tipsabout RNA and QPCR applications.

Contributed by Angie L. Bookout,Carolyn L. Cummins, andDavid J. Mangelsdorf

Howard Hughes Medical InstituteUniversity of Texas Southwestern

Medical CenterDallas, Texas

Jean M. Pesola and Martha F. Kramer(preparation of RNA standards)

Harvard Medical SchoolBoston, Massachusetts

Documents

High-Throughput Real-Time Quantitative · High-Throughput Real-Time Quantitative Reverse Transcription PCR 15.8.2 Supplement 73 Current Protocols in Molecular Biology amount of RNA